Laserfiche WebLink
no intense storms, despite higher average cover on <br />seeded plots. Higher dry season erosion was measured <br />on seeded plots the following year, which was attrib- <br />uted to pocket gophers attracted to the site by the <br />abundant ryegrass. Similar densities ofpocket gopher <br />mounds were found in operationally seeded areas <br />(Taskey and others 1989). <br />The moat extensive study of annual ryegrase effects <br />on erosion and vegetation response was conducted on <br />five sites burned in hot prescribed fires and awind- <br />drivenwildfire in coastal southern California (Beyera <br />and others 1998a, 1998b; Wohlgemuth and others <br />1998). Data on prefire vegetation and hillalope sedi- <br />ment movement were gathered, and greater replica- <br />tion was used than in moat previous studies. Fire <br />severity varied among sites from moderate to very <br />high, and poatfire precipitation varied from half of <br />normal to very high. Only plots that showed severity <br />effects great enough to trigger operational seeding <br />were retained in the study. At all five sites, poatfire <br />erosion was greatest duringthe firatyear after fire and <br />was not aignificantiy affected by ryegrass seeding <br />(Wohlgemuth and others 1998). Seeding increased <br />totalplantcoverthefirstyearatonlyonesite,byabout <br />1.5 percent, probably accounting for the lack of differ- <br />ence in erosion rates (Beyera and others 1998a). Aver- <br />age ryegrass cover reached 15 to 30 percent on some <br />sites during the second year after fire. Native herba- <br />ceous plant cover and species richness were lower on <br />seeded plots when ryegrasa cover was high (Beyera <br />and others 1994,1998b). Unlike some earlier studies, <br />Beyera and others (1998a) did not find significantly <br />lower shrub seedling density on seeded plots. In later <br />postfire years, some sites had significantly less ero- <br />sion on seeded than on unseeded plots, but this hap- <br />pened only after erosion rates had dropped to prefire <br />levels, which occurred in as little as 2 years on some <br />sites (Wohlgemuth and others 1998). Dry season ero- <br />sion (ravel) accounted for a high proportion of first- <br />year aedimentmovement on sites that burned during <br />early or mid summer. Grass seeding does not affect the <br />channel loading that occurs by this process (Beyera <br />and others 1998b, Wohlgemuth and others 1998). <br />These studies concluded that poatfire annual ryegrass <br />seeding is unlikely to reduce poatfire hillalope sedi- <br />ment movement the first year after fire in southern <br />California chaparral and has minimal impact on total <br />erosion from a bum site. <br />Grass species other than annual ryegrasa have <br />been used for postfire rehabilitation. Blando brome <br />($romus hordaceous cv "Blando°), promoted for use <br />in drought-prone areas, did not produce cover as well <br />as annual ryegrass (Blanford and Gunter 1972). <br />Canard and others (1995) tested several non-native <br />grasses and a native forb mix; only the native forb <br />mix significantly increased total plant cover, and <br />then only on anorth-facing elope. After the 1993 <br />firestorms in southern California, Keeley and others <br />(1995) found complete failure where native perennial <br />needlegraea (Nasella) species were used, and rela- <br />tively low levels of grass cover (1 to 23 percent) <br />produced bynon-native annuals such as Zorro fescue <br />( [~ulpia myuroscv "Zorro")and Blando brome, used to <br />avoid the competitive problems associated with an- <br />nual ryegrass. The highest seeded cover, 40 percent, <br />occurred on a Bite seeded with a mix of native species <br />and annual ryegrase. However, natural regeneration <br />ofnative and naturalized plants provided much more <br />cover than the seeded species. Although no direct <br />erosion measurements were made, Keeley and others <br />(1995) concluded that seeding was ineffective ae a <br />sediment control measure in the cases examined <br />because it contributed very little to total plant cover. <br />No quantitative studies on the impact of grass seed- <br />ing on poatfire erosion in chaparral have been pub- <br />lished from northern California or Arizona. Because <br />annual ryegrasa and other grasses typically establish <br />cover more successfully in northern California (Barra <br />and Canard 1987, Blanford and Gunter 1972), they <br />would be more likely to reduce erosion there. The <br />impact ofgraea seeding on native chaparral vegetation <br />in other areas, aside from suppression of shrub seed- <br />lings at very high grass densities (Schultz and others <br />1955), is largely unknown. <br />Conifer Forest: High intensity fire may be outside <br />the range of natural variability for many conifer plant <br />communities that have been subject to fire supprea- <br />sionfor the last century (Agee 1993). The lose of former <br />underatory seed banks due to overgrazing and canopy <br />denaification may also reduce the likelihood of rapid <br />regeneration of ground cover after fire. Seeding mixes <br />used in conifer stands often include legumes such ae <br />white clover (Trifolium repens) or yellow sweet clover <br />(Melilotus officinalis) to enhance nitrogen status of <br />the soil. Both annual and perennial grasses may be <br />used in mixes with non-native forage species origi- <br />nally tented for range improvement purposes (Christ <br />1934, Foraling 1931, McClure 1956), <br />Orr (1970) examined plant cover and erosion for <br />3 years after fire in the Black Hills of South Dakota <br />in an area operationally seeded with a mixture of <br />grasses and legumes. Most ofthe sediment production <br />occurred in two summer storms shortly after erosion- <br />measuring apparatus was set up. Sediment produc- <br />tion was inversely related to plant cover, Summer <br />storm runoff was 50 percentlesa on plate with high <br />plant and litter cover than on those with sparse cover. <br />Regression analysis showed that the decrease in run- <br />off and sediment production with increasing ground <br />cover leveled off at 60 percent cover, similar to results <br />presented by Noble (1966). Orr (1970) concluded that <br />seeded species were essential for quickly stabilizing <br />USDA Forest Service Gen. Tech. Rep. RMRS-GTR-63.2000 ~ ~ <br />